sophai/tsca

TSCA

Two-stage Clustering Algorithm

Copyright(c) 2016-, Po-Jen Hsu ([email protected])

Published under GNU General Public License Version 3, 29 June 2007

Author: Po-Jen Hsu

Requirements:

• Linux, Python(>=3.5), and numpy.

How to run:

1. Put tsca.py and tsca_pref.json in your working directory that contains xyz files.

2. Change the permission of tsca.py by typing chmod u+rx tsca.py.

3. Run python3 tsca.py tsca_pref.json to perform TSCA.

4. The resulting xyz files will be stored in the xyz folder by default.

Options in tsca_pref.json file:

1. "mode": "skip" will skip the job.

2. "processes": 1 will disable multiprocessing.

3. "output": "./some_path/" will specify the output directory.

4. "usr_threshold": float_value will assign the USR similarity threshold.

5. "pot_key": "eng=" will use "eng=" to parse the energy keyword in the info line of xyz files. Please refer to example1.xyz.

6. "use_cluster_id": false will disable the first stage clustering method that groups the ring/chain H-bond topologies.

Note:

1. Please delete the example*.xyz files before performing TSCA with your own xyz files. TSCA was designed to search for all the .xyz files in the current folder and the sub-folders.

2. You can also change the options from "filter_name:" ".xyz" to "filter_name": ".xyz -example" in the json file to exclude the example files.